Resume

Professional Experience

• Staff AI Researcher at Hootsuite (formerly Talkwalker), Luxembourg, August 2026 – present.
  
  • Design agents and skills, and build an evaluation framework combining deterministic rules and LLM-as-a-judge.
  • Co-develop the agent harness, including the orchestrator and memory layers.
• Quantitative Researcher at Abu Dhabi Investment Authority (ADIA), Abu Dhabi, July 2023 – July 2026.
  
  • Delivered features and signals used in live Market Neutral and Directional strategies traded daily in US and EU equity markets.
  • Co-founded and architected an NLP/LLM research platform for large-scale extraction of trading signals from unstructured financial text.
  • Designed and implemented sentence-granular pipelines for earnings calls, filings, and other unstructured data (parsing, splitting, labelling, embedding) to generate alpha features and signals for systematic strategies, with emphasis on point-in-time data.
• Senior Machine Learning Researcher at Talkwalker (now Hootsuite), Luxembourg, May 2021 – June 2023.
  
  • Owned the full lifecycle of NLP models, from initial research to production deployment, processing up to 10,000 documents per second.
  • Developed and scaled multilingual Transformer-based pipelines for aspect-based sentiment and emotion extraction from raw social media data.
  • Improved classifier accuracy in low-data regimes by implementing Active Learning and Meta-Learning research.
  • Optimized deep learning inference and data filtering logic to ensure high-fidelity signals across massive, unstructured data streams.
• Desk Strat at Deutsche Bank, London, September 2020 – April 2021.
  
  • Supported the Synthetic Desk (ETFs and Indices) migration to Kannon, the new P&L and Risk platform used by trading.
  • Improved PV and Risk computations for ETFs and indices.
• Research Advisor at ICA (now Opensee), Paris, May 2016 – March 2020.
  First as an intern, then as a PhD candidate, in collaboration with ICA:
  • Optimization of the analytic library using Optimal Quantization (pricing of exotic options in the interest rate world).
  • Identifying and killing bias in xVA computation using Multilevel Monte-Carlo methods.

Education

• PhD in Numerical Probability, Sorbonne University (LPSM), March 2017 – June 2020.
  Supervised by Gilles Pagès and Vincent Lemaire (LPSM) and Jean-Michel Fayolle (ICA).
  • Research on Optimal Quantization, Multilevel Monte-Carlo, and Stochastic Algorithms, spanning theory and financial applications.
  • Developed fixed-point and gradient-descent methods for efficient construction of optimal quantizers.
  • Applied the framework to PRDC Bermudan option pricing and the stationary Heston model.

• Research Master in Probability and Finance (Master El-Karoui, with honors), Sorbonne University in collaboration with École Polytechnique, September 2014 – June 2016.

• Bachelor’s degree in Mathematics (with honors), Aix-Marseille University, September 2011 – June 2014.

Skills

Programming

• Python (Pandas, Hugging Face, PyTorch, ONNX, NumPy, pytest, Pybind11, …)
• SQL (Redshift)
• C++ (libraries built during the PhD)
• Tools used often: Git, Airflow, S3
• Basic knowledge: Java, Scala, Kafka, Docker, gRPC, Elasticsearch

Languages

• French: native
• English: fluent